Author Topic: "Singular matrix" error in LTspice sim of Two transistor forward?  (Read 2048 times)

0 Members and 1 Guest are viewing this topic.

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Hi,
Do you know why this  LTspice sim of a two transistor forward wont run? The error message says “singular matrix node N029”.
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Re: "Singular matrix" error in LTspice sim of Two transistor forward?
« Reply #1 on: June 11, 2020, 08:07:04 pm »
Hi,
Yes I do.

Jay_Diddy_B
 
The following users thanked this post: T3sl4co1l, Ian.M, ocset

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Re: "Singular matrix" error in LTspice sim of Two transistor forward?
« Reply #2 on: June 12, 2020, 05:10:20 am »
Hi,

The singularity is caused by SG and FG only having one connection, B1 and B3 doesn't count.

Simply adding a resistor from SG and FG to ground will resolve this issue:



(Buffering the gate drive like that may have other bad effects, you have zero impedance in the gate drive circuit. I would connect SG and FG to the MOSFETs, like they are supposed to be.)

In addition you need to add UIC to the .tran statements.



These two modifications will allow the model to run.

Other Errors



Boost should not be connected to TS.



Remove R3, none of the datasheet schematic or demo boards have this resistor.



Remove the 10 \$\Omega\$ resistors.
The transformer turns ratio is backwards.

There may be more issues ...

NRND





I would look at the LTC3765/66 chipset.
Look for DC2199A-B

Regards,
Jay_Diddy_B


 
The following users thanked this post: ocset

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Re: "Singular matrix" error in LTspice sim of Two transistor forward?
« Reply #3 on: June 12, 2020, 10:00:04 am »
Thanks very much Jay_Diddy_B, greatly appreciated, i never thought i would see that sim come to life.
Its running now, but i think i need to tune the soft starts of LTC3705 and LTC3706 so the '3705 can hand over smoothly...also, my pulse transformer signal is noisy, and i need to look at the Ls and Cs in this Pulse transformer  circuit.

Incidentally i have it running fine with just the LTC3706 on its own (as attached).  I wonder why this kind of circuit isnt used...a digi isolator could port the drive signal across and the secondary side  sense resistor could be used as the current sense in current mode control
« Last Edit: June 12, 2020, 10:01:41 am by treez »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf